Full scale plant recovering iron phosphate from sewage at
Helsingborg Sweden
Ingemar Karlsson
Kemira Kemi AB, Kemwater
P.O.Box 902, S-251 09 Helsingborg
Sweden
ABSTRACT
Municipal wastewater sludge is a complex product containing both valuable products as well
as contaminants. The valuable products are phosphorous, organic matter and precipitants.
The contaminants include heavy metals and organic micro pollutants. Consumers, farmers
and their organisations are restrictive to utilise sludge on arable land.
Disposal of organic waste will be prohibited in 2005. Thus a new sludge treatment process,
KREPRO, has been developed. From the wastewater sludge different fractions can be separated and reused, i.e. organic sludge, phosphate, precipitants and carbon source.
The organic sludge has a dry solid content of about 50%, which makes it suitable for combustion. The ash content is about 30% of the dry solid matter. Test-firing of KREPRO sludge has
been performed in a steam boiler with moving grate. The sludge was co-fired with wood chip.
Co-firing of organic sludge and wood chip worked very well during the test. The boiler is
normally fired with different wood fuels.
The phosphate fraction is recovered as ferric phosphate with a very low heavy metal content.
About 75% of the phosphorus in the sludge can be recovered. Field tests have shown that the
phosphate is available for the crops and yield in the same order of magnitude as compared to
a commercial fertiliser. About 90% of the precipitant can be reused or recycled.
KEYWORDS
Sewage sludge, incineration, hydrolysis, phosphorus recycling
INTRODUCTION
All urban areas in Sweden are today connected to advanced sewage treatment with stringent
effluent standards. The sludge from sewage works is, if spread on farmland, regulated by
strict quality requirements set by the authorities. The content of heavy metals and organic
toxins is very low. But despite considerable interest from farmers, the spreading of sludge on
farmland is always questioned. The arguments put forward in the sludge debate are rarely
scientifically founded. They are marked more by fear and other emotional issues, which are
clearly sparked by a lack of knowledge about various known and unknown substances present
in the sludge and how they affect the environment. Lately there have been reports of contamination of silver and bromated flame-retardants. The next fright could be of hormones. The
only thing we know for certain is that there will be another alarm. The farmers’ union has
recommended their members not to spread the sewage sludge on farmland.
The processes today are designed to dilute or reduce the sludge. Dilution is normally made by
composting or mixing the sludge with lime. Reduction by hydrolysis improved dewatering or
drying. Disposal of organic waste will be prohibited by year 2005. The remaining options for
sludge will then be landfill or incineration. If sludge is disposed to landfill or incinerated, it is
no longer possible to recycle the phosphorus. Since new techniques for phosphorus recovery
are now being developed, the Swedish Environmental Protection Agency recommend in their
comment on the working document on sludge, that the new European Sludge Directive should
seek not only to promote various conventional ways of using sewage sludge, but also to open
the door for alternative ways of utilising phosphorus in sludge.
A research project on this subject, financially supported by the Swedish Environmental Protection Agency, was conducted during the 1990’. The aim of the project was to hydrolyse
sewage sludge and separating it into valuable recyclable components. The first full-scale application was taken into operation 1995 treating sewage sludge from 150 000 population
equivalents in south Sweden. The process was called ”KREPRO” (Kemwater Recycling
PROcess).
SEWAGE SLUDGE
Inside EU approximately 7,000,000 tons dry solid (DS) municipal sludge were produced in
1995. With increasing wastewater volumes and advanced wastewater treatment processes,
sludge production continues to grow. EU estimates that the sludge amount will be 12,000,000
tons DS per annum by 2005. [1]
Sludge from municipal sewage treatment plants has to some extent been spread on farmland
in Sweden. This is, however, a controversial subject. Farmers, consumers and food industry
do not accept sewage sludge as soil conditioner on lands for food and animal feed production.
Table 1 shows the limits for metals in sludge in Sweden, Germany and 86/278/EEC.
Table 1. Limits for metals in sludge for farming purpose in Sweden, Germany and
86/278/EEC 8.
Metal
Pb
Cd
Cu
Cr
Hg
Ni
Zn
Sweden,
mg/kg dry matter
BRD,
mg/kg dry matter
86/278/EEC, Annex IB,
mg/kg dry matter
100
2
600
100
2,5
50
800
900
10
800
900
8
200
2500
750-1200
20-40
1000-1750
16-25
300-400
2500-4000
A large percentage of the sludge produced today is deposited on landfills. However, long term
problems with leaching, heavy metal contamination and the solids content of the sludge, are
now becoming important issues. The number of areas available for sludge landfills is rapidly
decreasing, causing an increase of disposal costs. Landfill will therefore become a limited
option for disposal. In addition a disposal tax is in force in Sweden since year 2000.
THE KREPRO-PROCESS
The KREPRO sludge treatment process is able to separate valuable products from municipal
wastewater sludge. Four main products are being recovered from the sludge:
- Biofuel
- Phosphate
- Precipitant
- Carbon source
Both digested and raw sludge can be treated. The process is continuos and could be divided
into seven main steps:
- Thickening
- Acidification
- Thermal hydrolysis
- Biofuel separation
- Phosphorus precipitation
- Phosphate separation
- Recycling of precipitant and carbon source
The sludge is thickened to 5-7 % DS and acidified with sulphuric acid to pH between 1 and 3.
The coagulant, heavy metals and phosphorus are partly dissolved by this treatment. The organic suspended material is to a low degree solubilised.
The acidified sludge is heated to about 140OC in a pressure vessel. The retention time in the
reactor is 30-40 minutes and about 40% of the suspended organic matter are hydrolysed into a
biological readily degradable liquid. The inorganic compounds are now liquefied.
The undissolved organic matter, mainly fibres, now very easy to dewater, is separated in a
centrifuge to a dry solid content of about 50%. The volume reduction compared to conventional dewatered digested sludge is about 80%.
The energy content in this fraction is high, equal to that of wood chips, and sequentially it
may be used as a biofuel. The heavy metals can be separated together with the organic sludge
or later in the process.
The supernatant from the separation of organic sludge contains the inorganic substances from
which now the phosphorus is precipitated as a ferri phosphate.
This phosphate fraction is separated by centrifugation producing sludge of a dry solids content of 35%. The content of heavy metals and organic toxic substances is very low, why it
directly might be used as a fertiliser on farmland.
The liquid phase from the phosphate separation contains the precipitant, the dissolved organic
matter and nitrogen. This liquid fraction is recycled into the sewage plant for nutrient removal.
Figure 1. KREPRO – continuos process
Centrate
Acid
Inlet sludge
Digested or
Raw
2.5-4.0% TS
Thickener
centrifuge
Reactor
Flash tank
Mix tank
Steam
Fe
Heatexchanger
Alkali
Organic
centrifuge
Inorganic
centrifuge
Carbon source
and
recovery of
coagulant
FePO 4
35% TS
FePO 4
precipitation
Organic sludge
45% TS
Dewatered sludge with a dry solid concentration of about 20% DS can be treated in a batch
process. The energy is partly recovered by flushing. Because of the high concentration of the
sludge, the energy demand will be in the same order as for the KREPRO-process with heat
exchangers. The batch process enables a more compact treatment of the reject water.
Figure 2. KREPRO – batch process
External sludge
Acid
Vapour
Dilution
Digested
sludge
Mixin
DS
g 20 %
Dewatering
Reject water
treatment
Hydrolysi
s
Temp
150 oC
pH 1 - 2
pH adjustment
Ferric hydroxide
Heavy metals
separation
separation
Iron recovery
40 %
pH-adjustment
Sludge for
incineration
DS 45 %
Dewatering
pH adjustment
Phosphorus
precipitation
Iron phosphate
Phosphorus
recovery
75 %
ORGANIC SLUDGE PROPERTIES
Content
Organic sludge mainly consists of fibres and other undissolved organic matter. The dry solid
content in the organic sludge varies between 45 and 55%. The ash content is about 30% of dry
substance.
After removal of the nutrients, phosphorus and nitrogen in the organic fraction, the value for
agricultural use is minor. But the energy content is equal to that in wood fuels. The heavy
metals can be separated together with the organic sludge or later in the process.
Depending on the sludge quality used in the KREPRO-process and the pH during and after
the hydrolysis, there is a variation in the contents of the different substances within the organic sludge. Analyses of both the organic and the digested sludge have been conducted, and a
comparison of the restriction values on arable land was made. See Table 2.
Table 2. Content of heavy metals in sludge from the KREPRO-plant in Helsingborg,
compared to sludge restriction values for arable land in Sweden. [2,3]
Metal
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Digested sludge
mg/kg DS
1.7
35
450
45
1.0
30
550
Organic sludge
mg/kg DS
1.6
30
640
45
1.9
20
400
Limits in Sweden
mg/kg DS
2
100
600
100
2.5
50
800
Figure 3. Water, organic and inorganic fractions for different sludges
W ater, organic and inorganic fractions
in different sludges
Dewatering
1 ton TS
20% DS
Volume 5 m3
W ATER
O RG ANIC
INO RG ANIC
Drying
1 ton TS
90% DS
Volume 1,1 m3
W A T ER
KREPRO
0,5 ton TS
50% DS
Volume 1,0 m3
O R G A N IC
W A T ER
IN O R G A N IC
O R G A N IC
IN O R G A N IC
A typical analysis of the organic sludge is shown in table 3.
Table 3. Typical analysis of the organic sludge (4)
Dry solid matter
Ash content
C
H
O
N
S
Cl
P
%
% of dry solid matter
% of dry ash free matter
% of dry ash free matter
% of dry ash free matter
% of dry ash free matter
% of dry solid matter
% of dry solid matter
% of dry solid matter
45-55
15-30
49,7
6,8
33,8
2,1
2,5
<0,1
2,9
The nitrogen content in sludge is higher than in wood fuel with 0,3 % N of dry ash free matter. This may cause higher NOx-emissions. A great part of the sulphur in the sludge is present
as sulphate and will not contribute to SO2-emissons from the boiler.
The lower heat value for organic fuels depends on moisture content and ash content in the
fuel. The dry solid content in the organic KREPRO sludge varies between 45 and 55%. The
ash content in the dry substance is normally around 30 %. The heat value is in the range between 6-8 MJ/kg depending on dry solid content, which must be regarded as very high compared to untreated and dewatered municipal wastewater sludge.
The heat values of wet organic sludge, untreated digested sewage sludge and wood fuel are
shown in Table 4.
Table 4. The dry solid content, ash content and the lower heat value in
organic sludge (KREPRO), untreated digested sewage sludge and
wood fuel (branches and tops). (4)
Sample
Digested sewage sludge
KREPRO
Wood fuel
Wood fuel
Dry solid
content (%)
25
50
55
48
Ash (%) of dry
solid content
40
30
2,5
2,5
LHV
3,6
7,4
6,3
8,7
When the sludge is fired in a boiler, the ash composition is of interest. Based on the ash properties it is possible to predict risks of slagging and fouling. Preliminary findings show that the
risk of slagging is lower than for wood chip at furnace temperatures below 1200oC and
somewhat higher at higher temperatures. The risk of fouling is less when sludge is fired compared to wood chip.
Table 5 shows an analysis of ash composition of ash from KREPRO sludge.
Table 5. Ash composition, % (5)
SiO2
Al2O3
TiO2
Fe2O3
CaO
MgO
K2 O
Na2O
BaO
MnO
P2O5
SO3
%
61,4
15,2
2,2
3,8
1,3
1,6
2,8
3,2
1,0
0,04
5,4
2,0
TEST FIRING OF ORGANIC SLUDGE AND WOOD FUELS IN LINKÖPING
A combustion trial, where wood-chip, bark and organic sludge were mixed, was carried out at
a combined power and heating plant in Linköping, Sweden. The flue gases are cleaned from
dust, in sequence, in a cyclone separator, an electrostatic precipitator and then a condensing
plant. The trial was carried out with 58 tons of organic sludge from the KREPRO plant at
Helsingborgs STP. One part of the organic sludge was mixed with two parts bark and two
parts wood-chip.
The conclusion from the trial with organic sludge mixed with wood-chip and bark is that it is
possible to combust the different fuels together without any problems. The heat value in the
organic sludge is as high or higher than the one for wood chips and bark. No specific problems with the boiler when the organic sludge was combusted could be seen during the trial
PHOSPHORUS
The phosphorus recovered in the KREPRO-process is recovered as a ferric phosphate salt,
and has a low heavy metal content. Compared to an artificial fertiliser the level of heavy metals is the same or lower, see Figures 3 and 4. [6] 15% as P.
Figure 4. The content of heavy metals, expressed as mg HM/ kg phosphorus,
in ferric phosphate, digested sludge and artificial fertiliser; NPK 20:4:8
compared to the restriction values for sludge on arable land in Sweden.
90
80
70
mg/kg P
60
Cd
50
40
Hg
30
20
10
Artificial
fertiliser; NPK
20.4.8
SFS 1993:1271
§11
P=3%
Dewatered
digested sludge
dec-96
Öresundsverket
FePO4-970414
0
Figure 5. The content of heavy metals, expressed as mg HM/ kg phosphorus,
in ferric phosphate, digested sludge and artificial fertiliser; NPK 20:4:8
compared to the restriction values for sludge on arable land in Sweden.6
30000
25000
Zn
Ni
15000
Cr
10000
Pb
5000
Cu
Artificial
fertiliser; NPK
20.4.8
SFS 1993:1271
§11
P=3%
Dewatered
digested sludge
dec-96
Öresundsverket
0
FePO4-970414
mg/kg P
20000
Table 6. Heavy metals and some organic substances per kg P in digested sludge and
KREPRO phosphates
Heavy metals and some organic substances
per kg phosphorous
mg/kg P
Cu
Cd
Hg
Cr
Zn
Ni
Pb
Fluorine antene
Bensol(b)fluorine
PCB(52)
PCB(101)
Nonyl phenol
Toluene
Digested
Conventional Sludge
20 000
80
50
1 500
23 000
1 200
2 000
20
10
2,8
0,4
1 770
28
KREPRO Phosphate
100
3
1
220
1 000
300
180
<1
<1
24 x 10-3
<15 x 10-3
12
<5
Growth test with recovered phosphorus
Recovered phosphorous can either be reprocessed to different phosphorous products, or used
on arable land directly as a fertiliser. Through Kemwater’s KREPRO-process most of the
phosphorus is re-used as ferric phosphate. The availability of the phosphorus is important
when used as a fertiliser and is traditionally estimated by the solubility to various solutions..
EU directives, for instance specify several permissible solvents as a basis for evaluating phosphate firtilisers and their solubility: water where applicable, 2% formic acid, 2% citric acid,
Petermann´s and Joulie´s solutions as well as neutral ammonium citrate solutions. Petermann´s and Joulie´s solutions are alkaline ammonium citrates.
A test with water, 2% citric acid, pH 2 and ammonium citrate,pH 7 the solubility for
KREPRO ironphsphate and for different apatite based calcium phosphates is shown in table 7.
Ferric phosphate produced in the KREPRO-process is not water soluble but 100% ammonium
citrate soluble, which should be compared to an artificial fertiliser where the water solubility
ranges between 20-60% and the citrate solubility is 100%. The phosphate fraction, 35% TS,
contains about 10% P.
To establish whether it would be possible to use ferric phosphate directly in the fields as a
phosphorus source, the fertiliser was put to trial on ray grass during 7 months and 6 harvests.
These trials showed that the growth of the grass fertilised with ferric phosphate as the phosphorus source was as good as the one fertilised with artificial fertiliser. [7].Figure 6.
Table 7. Ammonium citrate, citric acid Solubilities of Calcium phosphates and Ferric phosphates
Ammonium citrate, citric acid Solubilities
of Calcium phosphates and Ferric phosphates
Calcium
phosphates
1. Kola apatite
2. Siilinjärvi
apatite
3. Jordan apatite
4. Ferric phosphate KREPRO
P2O5-total P2O5 water soluble
%
%
P2O5 ammonium
citrate soluble
%
P2O5 citric acid
soluble
%
38
<0,1
0,5
2,4
38
32
<0,1
<0,1
0,6
3,5
0,6
6,9
31
0,2
31
7,6
Figure 6. Evaluation of FePO4 as a fertilizer
Thanks to these results, more tests were conducted on different crops. During the spring and
summer of 1997 rape, pea and barley were tested in pot trials, with the latter being tested also
in field trials. These trials were performed and analysed by Svalöv Weibull (a plant breeding
company). [9]
It is possible to see the same trend in the pot trials as in the field trials. The only difference is
that the yield from the combination where no phosphorus was added was much lower.
Figure 7. Cultivation trails of barley
Cultivation trials of barley
16
Control
FePO4
-P
14
12
10
8
6
4
2
0
Total
plant
weight (g)
Seed
weight (g)
Ear per
plant
Seed per
ear
Conclusions
The findings from the field and pot trials, conducted by Svalöv Weibull, show that ferric
phosphate has a good effect when compared to the combination with no added phosphorus.
On the other hand it has a tendency to be 4-5% lower in yield compared to ordinary fertiliser.
A possible explanation to the lower yield might be that the plants, during the most intensive
growth period, have difficulties in utilising the phosphorus from the ferric phosphate. An
overdose of ferric phosphate would, however, increase the availability of phosphorus. This
investigation shows that the ferric phosphate has no negative effect on the flowering, the seed
or the seed growth. It also shows that similar results are to be expected in other crops.
THE CARBON SOURCE
When the organic fibre fraction, the phosphorus, the heavy metals, and the precipitant are
removed the reject water contains readily degradable organic matter and nitrogen. The
COD:N ratio is about 10:1 and thus very suitable for nitrogen removal in a Sequence Batch
Reactor SBR. The nitrogen in the reject water from dewatering of the sludge can also be treated in the SBR.
COSTS
Agriculture
Sewage sludge should be recycled to farmland if the quality demands are met. The value of
the sludge for agriculture is not only the phosphorous and nitrogen but also the organic content. The cost for dewatered sludge will be about 4 EURO per kg of P. A KREPRO process
can not complete and should not if other properties like cleaner sludge can not be shown.
Table 8. Sludge treatment costs Euro per ton DS
Sludge treatment costs Euro per ton DS
Conventional
dewatering 25% DS
Running
Investment
Handling
Total
KREPRO
25
15
85
100
100
20
125
220
Disposal
If sludge has to be dumped at waste dumps the reduction in volume with KREPRO makes the
total handling more feasible compared to conventional handling. The phosphorus content will
be lost if it is no removal before disposal.
Table 9. Sludge treatment costs Euro per ton DS
Sludge treatment costs Euro per ton DS
Disposal
Conventional
dewatering 25% DS
KREPRO
125
220
Fee (30)
Tax (30)
120
120
30
30
Total
365
280
Incineration
If sludge is incinerated together with garbage a KREPRO process is much more feasible, table
10, compared to conventional dewatered sludge. Dewatered sludge 25% DS has a negative
energy value when incinerated, KREPRO a positive. In the KREPRO process the organic
sludge has an energy content equal to wood chips and the phosphorous is recovered before
incineration. The volume of a KREPRO sludge is reduced with about 75% compared to a
conventional dewatered sludge, or the same volume as a dried sludge.
Table 10. Sludge treatment costs Euro per ton DS Incineration
Sludge treatment costs Euro per ton DS
Incineration
Conventional
dewatering
25% DS
KREPRO
125
220
Fee (70)
Disposal ash
Fee (30)
Tax (30)
280
70
15
15
4
4
Total
435
298
CONCLUSIONS
The KREPRO-process separates valuable products from municipal wastewater sludge such as
phosphate, precipitant, carbon source and organic sludge. The organic sludge has a high content of dry matter compared to untreated sludge and is suitable for combustion.
The test firing showed that it was possible to co-fire KREPRO sludge and wood chip in a
boiler with a moving grate without problems. Another conclusion was that the KREPRO
sludge could be fired without investments in separate sludge feeding equipment.
The phosphate fraction is recovered as ferric phosphate with a very low heavy metal content.
About 75% of the phosphorus in the sludge can be recovered. Field tests have shown that the
phosphate is available for the crops and the yield is in the same order of magnitude as compared to a commercial fertiliser. About 90% of the precipitant can be re-used or recycled.
The volume reduction of the libre fraction compared to dewatered sludge makes the
KREPRO-process lower in total cost if the sludge is deposited or incinerated.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
A. King: Global water supply & sewerage 1997. Water & Environmental International
SNFS, 1994 Kunggörelse SNFS 1994:2
Svenska Naturvårdsverket: Fosfor och energi ur avloppsslam. Rapport 4822
Faria, P. Exam work. Lunds Technical University. Användning av slam som bränsle
T. Öhrström: Provförbränning av behandlat rötslam i panna 3 KKV, Linköping. Rapport TÖh-97/21,
Bergström & Öberg.
Cassidy, S. (1996) Recovery of valuable products from municipal waste water sludge. Chemical Water and
Wastewater, H. H. Hahn, E. Hoffman and H. Ødegaard (Eds) © Springer-Verlag Berlin Heidelberg pp. 325-340
M. Koivunen: Evaluation of FePO4 as a fertiliser- ERC pot trial 1996. Kemira Agro OY. 96034-Esp-Re
Eksvärd, J. LRF. Personal communication
N-O. Bertholdsson: FePO4 som fosforgödning till korn, ärt och raps. Bilaga till Naturvårdsverket